So I asked myself the question, how does "big pharma" develop drug new drug candidates.
That's when I learned about Docking, I also came across software for drug design.
http://en.wikipedia.org/wiki/Docking_%28molecular%29
http://www.imb-jena.de/~rake/Bioinformatics_WEB/dd_tools.html
http://www.click2drug.org/
Now lately I've been reading about the various serotonin receptors, and think it would be a cool idea to create an open source program that uses a combination of reference molecular structures + Ki nm, docking and a neural network to look for rules in structures to generate molecular candidates. Specifically geared to the human 5ht receptors.
Imagine being able to enter in various 5ht receptor affinities and the program calculates several candidates that are the closest to the parameters given.
Also in my recent studies of 5ht receptors I have learned why those SSRAs do not feel like mdma(mainly missing that 5ht2b agonism that causes the body high and burst of serotonin), why mescaline makes people puke(5ht3 agonism) and why lsd causes weird effects on memory(5ht6 agonism).
Anyone up for starting this project?
Also what would a set of parameters would you give the software to work with?
For example, mine would be:
5ht1a agonist ~1.3 Ki
5ht2 a ~0.89 Ki and b ~3.3 Ki agonists c antagonist ~6 Ki
5ht6 antagonist ~6.5 Ki
5ht7 agonist ~1.94 Ki
That's when I learned about Docking, I also came across software for drug design.
http://en.wikipedia.org/wiki/Docking_%28molecular%29
http://www.imb-jena.de/~rake/Bioinformatics_WEB/dd_tools.html
http://www.click2drug.org/
Now lately I've been reading about the various serotonin receptors, and think it would be a cool idea to create an open source program that uses a combination of reference molecular structures + Ki nm, docking and a neural network to look for rules in structures to generate molecular candidates. Specifically geared to the human 5ht receptors.
Imagine being able to enter in various 5ht receptor affinities and the program calculates several candidates that are the closest to the parameters given.
Also in my recent studies of 5ht receptors I have learned why those SSRAs do not feel like mdma(mainly missing that 5ht2b agonism that causes the body high and burst of serotonin), why mescaline makes people puke(5ht3 agonism) and why lsd causes weird effects on memory(5ht6 agonism).
Anyone up for starting this project?
Also what would a set of parameters would you give the software to work with?
For example, mine would be:
5ht1a agonist ~1.3 Ki
5ht2 a ~0.89 Ki and b ~3.3 Ki agonists c antagonist ~6 Ki
5ht6 antagonist ~6.5 Ki
5ht7 agonist ~1.94 Ki